Unverified50% confidenceFactExact time
KeyASRBench proposes an occurrence-based Keyword Error Rate (KER), computed on frozen whole-phrase annotations, treating each keyword occurrence as an independent evaluation unit and counting it as preserved only if the entire phrase is completely correct.
1
Sources
50%
Confidence
Long-term
Relevance
11/10/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedWER treats all word errors equally, applying the same penalty regardless of whether the error involves a function word or a key entity word.64% similarUnverifiedEach word output by Tesseract comes with a confidence score, which can be used to flag low-confidence words for human review.63% similarVerifiedWord Error Rate (WER) is calculated by dividing the sum of substitution, deletion, and insertion errors between the recognition output and the reference text by the total number of words in the reference text.63% similarUnverifiedCharacter Error Rate (CER) is defined as the edit distance (insertions + deletions + substitutions) divided by the total number of characters in the reference text; Word Error Rate (WER) is analogous but operates on words.62% similarUnverifiedThe typical KE pipeline includes candidate generation, feature construction, ranking or classification, and final keyword annotation.61% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1153736API
curl https://kongchang.com/api/v1/knowledge/claims/1153736MCP
get_claim(id=1153736)